PERIOD2 is a circadian negative regulator of PAI-1 gene expression in mice

Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

Contents lists available at ScienceDirect

Journal of Molecular and Cellular Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y j m c c

Original article

PERIOD2 is a circadian negative regulator of PAI-1 gene expression in mice Katsutaka Oishi a,⁎, Koyomi Miyazaki a, Daisuke Uchida a,b, Naoki Ohkura c, Miyuki Wakabayashi a, Ryosuke Doi a,b, Juzo Matsuda c, Norio Ishida a,b,⁎ a Clock Cell Biology Research Group, Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology, Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan b Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8502, Japan c Department of Clinical Molecular Biology, Faculty of Pharmaceutical Sciences, Teikyo University, 1091-1 Suarashi, Sagamiko, Sagamihara, Kanagawa 229-0101, Japan

a r t i c l e

i n f o

Article history: Received 16 October 2008 Received in revised form 5 January 2009 Accepted 6 January 2009 Available online 10 January 2009 Keywords: Plasminogen activator inhibitor-1 Circadian clock Obesity Period2 E-box Nuclear localization domain BMAL1

a b s t r a c t An increased level of obesity-induced plasma plasminogen activator inhibitor-1 (PAI-1) is considered a risk factor for cardiovascular disease. To determine whether the circadian clock component PERIOD2 (PER2) is involved in the regulation of PAI-1 gene expression, we performed transient transfection assays in vitro, and generated transgenic (Tg) mice overexpressing PER2. We then compared PAI-1 expression in Tg and wildtype (WT) mice with or without obesity induced by a high-fat/high-sucrose diet. PER2 suppressed CLOCK: BMAL1- and CLOCK:BMAL2-dependent transactivation of the PAI-1 promoter in vitro. Furthermore, nuclear translocation is dispensable for PER2 to suppress CLOCK:BMAL1-dependent transactivation of the PAI-1 promoter, because functional loss of the nuclear localization domain did not affect either the interaction with BMAL1 or the suppressive role of PER2. The diurnal expression of clock and clock-controlled genes was disrupted in a gene-specific manner, whereas that of PAI-1 mRNA was significantly damped in the hearts of PER2 Tg mice fed with a normal diet. Obesity-induced plasma PAI-1 increase was significantly suppressed in Tg mice in accordance with cardiac PAI-1 mRNA levels, whereas body weight gain and changes in metabolic parameters were identical between WT and Tg mice. Endogenous PAI-1 gene expression induced by transforming growth factor-β1 was significantly attenuated in embryonic fibroblasts derived from Tg mice compared with those from WT mice. Our results demonstrated that PER2 represses PAI-1 gene transcription in a BMAL1/2-dependent manner. The present findings also suggest that PER2 attenuates obesity-induced hypofibrinolysis by downregulating PAI-1 expression independently of metabolic disorders. © 2009 Elsevier Inc. All rights reserved.

1. Introduction Obesity is an independent risk factor for the development of atherosclerosis and cardiovascular disease [1,2]. The inhibition of fibrinolysis and obesity are closely connected, and elevated levels of plasma plasminogen activator inhibitor-1 (PAI-1), the primary physiological inhibitor of plasminogen activators, are contribute to various vascular pathologies including coronary artery thrombosis [1–4]. Various cell types in vitro produce PAI-1, which is widely distributed in tissues such as vessel walls (endothelial and smooth muscle cells), the liver and adipose tissues as well as in macrophages. The expression of PAI-1is regulated by several cytokines, hormones and metabolic factors such as tumor necrosis factor-α (TNF-α), transforming growth factor-β1 (TGF-β1), insulin, glucocorticoids, angiotensin II, some fatty acids, and glucose [5,6].

⁎ Corresponding authors. Clock Cell Biology Research Group, Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology, Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan. Katsutaka Oishi is to be contacted at Tel.: +81 29 861 6053; fax: +81 29 861 9499. E-mail addresses: [email protected] (K. Oishi), [email protected] (N. Ishida). 0022-2828/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.yjmcc.2009.01.001

Serious adverse cardiovascular events including myocardial infarction, sudden cardiac death, pulmonary embolism, critical limb ischemia and aortic aneurysm rupture, all have pronounced circadian rhythmicity that peaks during the morning [7–9]. The frequency of myocardial infarction during this period is 1.5- to 3fold higher than that at other times of the day. Levels of blood PAI-1 peak during the early morning, which might explain the morning onset of myocardial infarctions [10,11]. Maemura et al. [12] described the circadian expression of PAI-1 mRNA in the heart and kidneys of mice, and suggested that the circadian oscillation of PAI-1 gene expression plays an important role in the circadian fluctuation of blood fibrinolytic activity. Assays in vitro have shown that CLOCK: BMAL2 (CLIF) and CLOCK:BMAL1 heterodimers up-regulate human PAI-1 gene expression via E-box (CACGTG) elements located at bp −677 to −672 and at bp −562 to −557 [12,13]. The core component of the circadian clock, REV-ERBα, is also involved in circadian regulation of the human PAI-1 gene via two ROR/REV-ERB binding elements (ROREs) in the promoter [14]. We recently demonstrated that the circadian clock proteins CLOCK and BMAL1 effectively transactivate the mouse PAI-1 gene [15], as well as that of humans [12,13].

546

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

Basic helix–loop–helix (bHLH)-PAS transcription factors such as CLOCK and BMAL1 are positive regulators of an autoregulatory transcription-translation feedback loop of the molecular circadian clock [16]. Both CLOCK and BMAL1 transactivate other clock genes such as period1 (Per1), Per2, cryptochrome1 (Cry1) and Cry2 via E-box elements in their promoters [16,17]. The PER and CRY proteins are translocated to the nucleus as multimeric complexes. The CRY proteins are essential for the negative feedback loop that regulates the central clock [18,19]. The primary function of CRY proteins in mammals is to inhibit CLOCK/BMAL1-mediated transactivation [16,17]. Per2 is also a core component of the circadian clock because the circadian period in mice with a Per2 mutation is shorter and it is followed by a complete loss of circadian behavioral activity in constant darkness [18]. On the other hand, the functional role of PER2 protein in molecular clock regulation is not fully understood, although PER2 is involved in various physiological functions such as tumor suppression [20,21], angiogenesis [22], immune function [23] and aortic vascular endothelial function [24]. Although PER2 overexpression modestly suppresses CLOCK:BMAL1-dependent transactivation in vitro [19,25,26], the transcription of CLOCK:BMAL1dependent circadian genes such as Per1 and Cry1 is positively regulated by PER2 in vivo [27,28]. We previously showed that the circadian expression of PAI-1 mRNA is damped in Clock mutant mice [29]. Furthermore, a Clock mutation attenuates the plasma PAI-1 increase in accordance with PAI-1 mRNA expression induced by metabolic disorders such as diabetes and obesity [30,31]. CLOCK is involved in glucose and lipid metabolism [32–37]. In fact, we demonstrated that the Clock mutation suppresses obesity-induced hyperglycemia and hyperinsulinemia independently of increasing body weight (BW) and visceral fat accumulation in mice [31]. Therefore, we could not determine whether the Clock mutation directly or indirectly suppresses diabetesand obesity-induced PAI-1 gene expression [30,31]. Here, we showed that PER2 plays a suppressive role in CLOCK: BMAL1 and CLOCK:BMAL2-induced PAI-1 gene expression in vitro. We also generated transgenic mice overexpressing PER2 and examined PAI-1 expression under diet-induced obesity to assess the function of PER2 in the circadian regulation of fibrinolysis and of obesity-induced hypofibrinolysis in vivo.

membranes (Bio-Rad), and then non-specific binding was blocked with 3% dried milk in 0.05% Tween 20-PBS. The membranes were briefly washed and then immunoblotted against anti-rPER2 antiserum, antiFLAG M2 antibody (Sigma) or anti-Clock (S-19) antibody (Santa Cruz Biotechnology, Inc.). Immunoreactive bands were visualized using horseradish peroxidase-conjugated secondary antibodies and the ECL detection system (Amersham Biosciences). 2.3. Animals and diet Transgenic mice over-expressing rat PER2 were generated by excising rPer2 cDNA (bases 1–3897; GenBank accession number AB016532) using the C57BL/6J strain [40]. The cDNA was ligated downstream of the constitutive cytomegalovirus (CMV)-IE enhancer/ β-actin promoter. Male wild-type (WT) and PER2 Tg mice at 4 weeks of age were maintained under a 12:12 h light–dark cycle (lights on at 0:00 and lights off at 12:00) and fed with a normal diet (ND) (CE-2; Clea Japan Inc., Tokyo, Japan) or with a high-fat/high-sucrose diet (HFSD) (F2HFHSD; Oriental Yeast Co. Ltd., Tokyo, Japan) containing 54.5% fat, 28.3% carbohydrates and 17.2% protein for 12 weeks to induce obesity. The mice were then sacrificed and tissues were dissected, quickly frozen and stored in liquid nitrogen. 2.4. Semi-quantitative reverse transcription (RT)-PCR Total RNA was extracted using guanidinium thiocyanate followed by RNAiso (Takara Bio Inc., Otsu, Japan) and digested with DNase I (Applied Biosystems/Ambion, TX, USA). Single-strand cDNA was synthesized using the PrimeScript™ RT reagent kit (Takara Bio Inc., Otsu, Japan). Real-time RT-PCR was conducted with the SYBR® Premix Ex Taq™ II (Takara Bio Inc., Otsu, Japan) using a LightCycler™ (Roche Diagnostics, Mannheim, Germany). The reaction conditions were 95 °C for 10 s followed by 45 cycles of 95 °C for 5 s, 57 °C for 10 s and 72 °C for 10 s. Table 1 shows the sequences of the primer pairs. The amount of mRNA was corrected relative to that of β-actin. 2.5. Western blots for tissue PER2 Mouse tissues were homogenized in tissue lysis buffer [31]. Total protein (150 μg) was analyzed by SDS-PAGE on 7.5% gels as described

2. Materials and methods 2.1. Transient transfection assays Mouse NIH3T3 cells were cultured and transient transfection assays were performed as described [15]. Mouse CLOCK, BMAL1, and CRY1 expression plasmids were provided by Dr. T. Todo [38]. The BMAL2 expression plasmid was provided by Dr. M. Ikeda. Expression vectors encoding rat intact PER2 and nuclear localization domain (NLD)-deleted (NLD(−)) PER2 have been described [39]. 2.2. Immunoprecipitation and Western blots Mouse BMAL1 and Clock cDNAs were ligated into pFLAG-CMV2 (Sigma) and pcDNA3.1-His (Invitrogen) plasmids, respectively. COS-1 cells were cultured as described [15]. Intact rPER2 or NLD(−) PER2 were cotransfected with mBMAL1 and mCLOCK in COS-1 cells. Twenty-four hours later, transfected COS-1 cells were washed twice with PBS, lysed with cell lysis buffer (150 mM NaCl, 5 mM EDTA, 0.5% NP-40 and 50 mM Tris–HCl [pH 7.5]) including protease inhibitor cocktail (Complete™; Roche Diagnostics), and incubated for 30 min on ice. Cell lysates prepared by centrifugation at 15,000 g for 10 min at 4 °C were immunoprecipitated using anti-FLAG M2 agarose beads (Santa Cruz Biotechnology, Inc.), which were then washed three times for 10 min each at 4 °C and boiled in 2× SDS sample buffer. The immunoprecipitated proteins were separated on SDS-PAGE and transferred onto nitrocellulose

Table 1 Primer sequences for quantitative real-time PCR Gene

Direction

Primer sequence

Per2

Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse

CACTCAGGAGTGCATGGAGGAGA CTGCTCTTGCACCTTGACCAGGT CGCCATCGACGTAACAGGG ATTGGCCAGCTGTCCGCTC TTCCACTATGTGACAGCGGAGG CGTATCCATTCATGTCGGGCTC CCAGATTGGTGGAGGTTACTGAGT GCGAGAGTCTTCTTGGAGCAGTAG CCCTGGACTCCAATAACAACACA GCCATTGGAGCTGTCACTGTAG AGGAGGACAGATCCCAATGGA GCAACCTTCTGGATGCCTTCT GAAGCACGTGAAAGCATTGACA CCCGACAAATCACCAGCTTG GGACACCCTCAGCATGTTCA TCTGATGAGTTCAGCATCCAAGA GGAACTGAAGCCTCAACCAAT CTCCGGCTCCAGTACTTCTCA AGCACAACAGCTGACTACGATAAG GCGCTTCCGGCACGCTGGAATGATCTAA CACACCTTCTACAATGAGCTGC CATGATCTGGGTCATCTTTTCA

rPer2 Tg mPer2 Per1 Rev-erbα Cry1 Dec1 PAI-1 DBP MMP-9 β-actin

Per2, total Per2 (endogenous mPer2 and exogenous rPer2); rPer2 Tg, exogenous rPer2; mPer2, endogenous mPer2.

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

547

2.7. Measurement of plasma and tissue PAI-1 Plasma and tissue PAI-1 levels were measured as described [31]. 2.8. Cell culture and stimulation Mouse embryonic fibroblasts (MEF) [42] derived from WT and PER2 Tg mice were cultured as described [35]. At 0 h, the cells were stimulated with 10 ng/ml TGF-β1 (Sigma), then washed with ice-cold PBS at the indicated times, harvested in 1 ml of RNAiso and stored at −80 °C. 2.9. Statistics Data are expressed as means ± SD or as means ± SEM. Group variations were statistically analyzed using the one-way or two-way analysis of variance (ANOVA) and further tested by Student's or Welch's t-test with P b 0.05 as the criterion for statistical significance. 3. Results Transient transfection assays of NIH3T3 cells showed that CLOCK: BMAL1 and CLOCK:BMAL2 heterodimers can activate the mouse PAI1-promoter in an E-box-dependent manner (Fig. 1(A)). Cotransfected PER2 suppressed CLOCK:BMAL1- and CLOCK:BMAL2-induced PAI-1promoter activation by 44% and 47%, respectively, whereas cotransfected CRY1 completely inhibited both activations (Figs. 1(B) and (C)). To determine whether nuclear translocation is essential for the suppressive role of PER2, we cotransfected the cells with PER2 lacking a functional NLD (Figs. 1(B) and (C)). Most of the overexpressed NLD(−) PER2 was located in the cytosol (data not shown) as shown in COS-1 cells [39]. Thus, the lack of an NLD did not affect the suppressive effect of PER2 on CLOCK:BMAL1-induced PAI-1promoter activation (Fig. 1(B)), although NLD(−) PER2 could not suppress the transactivity of CLOCK:BMAL2 (Fig. 1(C)). To determine whether the NLD is required for PER2 to bind with BMAL1, intact or NLD(−) PER2 was coexpressed with CLOCK and FLAGtagged BMAL1 in COS-1 cells, and then immunoprecipitated with antiFLAG antibody (Fig. 2). Phosphorylated and unphosphorylated forms

Fig. 1. PER2 differentially suppresses CLOCK:BMAL1- and CLOCK:BMAL2-dependent transactivation of PAI-1-promoter activity. (A) NIH3T3 cells were transfected with a reporter plasmid containing intact or E-box-deleted promoter region of mouse PAI-1 gene. (B, C) Effect of overexpression of intact PER2 or NLD(−) PER2 on PAI-1-promoter activity induced by (B) CLOCK:BMAL1 or (C) CLOCK:BMAL2 heterodimer. Values represent means ± SD (n = 4) and are representative of 3 to 5 independent experiments. Different characters indicate statistical significance (P b 0.05).

above. Anti-mouse PER2 antiserum was raised in a rabbit to GST (glutathione S-transferase) fused to mKIAA antigen GX0265 that includes the mPER2 carboxyl terminal region (amino acids 1105– 1257) [41]. The antiserum recognized overexpressed rPER2 as well as mPER2 (data not shown). 2.6. Measurement of blood metabolic parameters Platelet-poor plasma was collected and stored as described [31]. Plasma glucose, triglyceride (TG), total cholesterol (T-Cho), and insulin levels were measured as described [31].

Fig. 2. Nuclear localization domain (NLD) of PER2 is not required for interaction with BMAL1. Intact rPER2 or NLD(−) PER2 were cotransfected with mBMAL1 and mCLOCK in COS-1 cells. Cell lysates were immunoprecipitated (IP) with anti-FLAG M2 agarose beads, separated on SDS-PAGE, and immunoblotted against anti-rPER2 antiserum, antiFLAG M2 antibody or anti-CLOCK antibody. Input shows 17% of each cell lysate.

548

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

of CLOCK were coimmunoprecipitated with BMAL1 as described [43]. We found that NLD(−) PER2 can interact as well as intact PER2 with BMAL1. Immunoprecipitation with anti-rPER2 antibody also confirmed interactions between intact or NLD(−) PER2 and BMAL1 (data not shown). Mice overexpressing rat PER2 were generated as described [40]. The rat transgene is highly homologous to the mouse gene and thus mRNA from the transgene is easily distinguishable from endogenous mouse loci. The behavioral rhythms of the PER2 Tg mice are normal under LD, although the free-running period is shorter by 0.7 h under constant darkness [40]. To evaluate the effect of PER2 overexpression on the cardiac clock, we examined the expression profiles of clock and clock-controlled genes in the hearts of Tg mice (Fig. 3). The mRNA

Fig. 3. Temporal cardiac expression profiles of circadian and E-box-dependent clockcontrolled genes in PER2 Tg mice. Mice were maintained under a 12:12 h light–dark cycle (lights on at 0:00 and lights off at 12:00). Total RNA was extracted from hearts of wild-type (WT, open circles) and PER2 Tg (closed circles) mice, and mRNA levels were quantified by RT-PCR. Maximal value for WT mice is expressed as 100% except for rPer2. Values are means ± SD (n = 3). Significant differences compared with values from WT mice at each time point are indicated as ⁎P b 0.05. Per2, total Per2 (endogenous mPer2 and exogenous rPer2); rPer2 Tg, exogenous rPer2; mPer2, endogenous mPer2.

Fig. 4. Protein expression levels of PER2 and PAI-1 in hearts of PER2 Tg mice. Mice were maintained under a 12:12 h light–dark cycle (lights on at 0:00 and lights off at 12:00) and dissected at 14:00. Representative Western blots of PER2 (mPER2 and overexpressed rPER2) in hearts of wild-type (WT) and PER2 Tg mice (A). Arrowhead, position of PER2. Graphs show PER2 and PAI-1 expression levels determined by Western blotting (A) and ELISA (B), respectively. Open and closed bars indicate WT and PER2 Tg mice, respectively. Mean value for WT mice is expressed as 100% for (A). Values are means ± SEM (n = 3–5). Significant differences compared with value from WT mice are indicated as ⁎P b 0.05.

Fig. 5. Body weight (BW) gain and blood metabolic parameters in PER2 Tg mice fed with normal (ND) or high-fat/high-sucrose (HFSD) diets. (A) Gains in BW. Values are means ± SEM (n = 7–15). (B) Plasma metabolic parameters. Blood plasma was collected from wild-type (WT) and PER2 Tg mice at 2:00 (open bars) and 14:00 (closed bars). Values are means ± SEM (n = 5). Significant differences compared with value of ND for each genotype are indicated as #P b 0.05.

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

549

Fig. 7. Plasma total PAI-1 levels in PER2 Tg mice fed with normal (ND) or high-fat/highsucrose (HFSD) diets. Blood plasma was collected from wild-type (WT) and PER2 Tg mice at 2:00 (open bars) and 14:00 (closed bars). Values are means ± SEM (n = 5). Significant difference compared with value of WT mice at each time point is indicated as ⁎P b 0.05. Significant differences compared with value of ND in each genotype are indicated as #P b 0.05. Significant day/night fluctuations in each genotype are indicated as †P b 0.05.

Twelve weeks of feeding with the HFSD significantly increased BW by about 60% in both WT and Tg mice compared with mice on the ND (Fig. 5(A)). The increase in BW between WT and Tg mice did not significantly differ. Plasma T-Cho and insulin levels were increased by HFSD in both WT and Tg mice, whereas the plasma glucose and TG levels were not affected by HFSD in both genotypes (Fig. 5(B)). The insulin-to-glucose ratio (a measure of insulin resistance [47,48]) was significantly increased in both genotypes (Supplemental Fig. 2), although the plasma glucose levels were not significantly affected by HFSD (Fig. 5(B)). Glucose tolerance did not statistically differ between WT and Tg mice with HFSD-induced obesity (Supplemental Fig. 3).

Fig. 6. Expression profiles of Per2 and PAI-1 mRNAs in PER2 Tg mice fed with normal (ND) or high-fat/high-sucrose (HFSD) diets. Total RNA was extracted from hearts, livers and epididymal adipose tissues at 2:00 (open bars) and 14:00 (closed bars). Value at 14:00 of wild-type mice fed with ND is expressed as 100% in each tissue. Values are means ± SEM (n = 5). Significant differences compared with value of WT mice at each time point are indicated as ⁎P b 0.05. Significant differences compared with value of ND in each genotype are indicated as #P b 0.05. Significant day/night fluctuations in each genotype are indicated as †P b 0.05.

expression levels of total Per2 (endogenous mPer2 and exogenous rPer2) rather surprisingly fluctuated in a diurnal manner in the hearts of PER2 Tg mice. Furthermore, the diurnal rhythm of exogenous rPer2 gene expression was robust in phase with endogenous mPer2 expression, although the transgene does not possess the E-box enhancer that is responsible for circadian transactivation of the Per2 gene [44]. We found that PER2 overexpression obviously damped diurnal expression of the PAI-1 gene as well as of other circadian genes such as Per1, Rev-erbα, and DBP without affecting endogenous mPer2 and Dec1 expression (Fig. 3), although all of these genes are transactivated by CLOCK/BMAL1 via the E-box in a circadian manner [12,15– 17,45,46]. Fig. 4 shows that PER2 was overexpressed at the protein level and that PAI-1 protein levels were decreased in the hearts of Tg mice. Levels of Cry1 mRNA were continuously elevated in PER2 Tg mice, although the circadian expression of Cry1 is also regulated by CLOCK/BMAL1 [19]. Expression levels of transcription factors such as CLOCK, NPAS2, BMAL1 and BMAL2 that are potentially involved in transactivation of the PAI-1 gene were increased or identical in Tg mice compared with those in WT mice (Supplemental Fig. 1).

Fig. 8. TGF-β1-induced PAI-1 gene expression is attenuated in PER2 Tg mice in vitro. Mouse embryonic fibroblasts (MEF) derived from wild-type (WT) and PER2 Tg mice were stimulated with TGF-β1. Expression levels of PAI-1 and matrix metalloproteinase-9 (MMP-9) mRNAs were evaluated at indicated times. Open and filled circles indicate values for WT and PER2 Tg MEF, respectively. Pre-stimulation value of WT MEF is expressed as 100%. Values are means ± SD (n = 3). Significant differences compared with value of WT MEF at each time point are indicated as ⁎P b 0.05.

550

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

We found that Per2 mRNA was overexpressed in the heart, liver and adipose tissues of Tg mice under both ND and HFSD, but most remarkably in the heart (Fig. 6). Day/night fluctuation of Per2 mRNA expression was slightly and severely damped by HFSD-induced obesity in tissues of WT and Tg mice, respectively, and PAI-1 mRNA expression also similarly fluctuated in tissues from WT and Tg mice under ND. The PAI-1 mRNA levels were extremely elevated in these tissues from WT mice with HFSD-induced obesity. The augmented PAI-1 mRNA expression levels were identical between WT and Tg mice in the liver and adipose tissues. However, cardiac PAI-1 mRNA levels were significantly suppressed in obese Tg mice like those under ND, and the expression levels in Tg mice on the HFSD were lower than those in WT mice on the ND. Plasma PAI-1 levels fluctuated in a diurnal manner in both WT and Tg mice, and did not significantly differ between the genotypes under ND (Fig. 7). Plasma PAI-1 levels were remarkably increased by the HFSD in both genotypes. Notably, the nocturnal increase in plasma PAI-1 levels was significantly attenuated in Tg mice fed with the HFSD. To determine the functional role of overexpressed PER2 protein in Tg mice at the cellular level, embryonic fibroblasts from WT and Tg mice were stimulated with TGF-β1, which is a powerful inducer of PAI-1 gene expression (Fig. 8). Levels of PAI-1 mRNA were slightly but significantly lower in unstimulated MEF from PER2 Tg than from WT mice. The expression of PAI-1 mRNA was obviously induced by TGF-β1 in MEF from WT mice, whereas the induction of PAI-1 gene expression was attenuated in MEF from PER2 Tg mice. On the other hand, TGF-β1-induced matrix metalloproteinase-9 (MMP-9) expression was clearly augmented in MEF from Tg, than from WT mice. 4. Discussion We showed here that PER2, a core component of the mammalian circadian clock, is involved in the circadian regulation of PAI-1 gene expression both in vitro and in vivo. Transient transfection assays showed that CLOCK:BMAL1- and CLOCK:BMAL2-dependent activation of the PAI-1-promoter was significantly suppressed by PER2, although CRY1 seemed to be a stronger repressor than PER2. Thus, PER2 and CRY1 apparently inhibit CLOCK:BMAL1/2-dependent transactivation in different ways [19,49]. Our findings indicated that nuclear entry is dispensable for PER2 to inhibit CLOCK:BMAL1-dependent transactivation of the PAI-1 gene, although nuclear translocation is indispensable for PER2 to inhibit CLOCK:BMAL2-dependent transactivation. Interaction has been reported between PER2 and BMAL1 both in vitro [49] and in vivo [43,50]. As the NLD(−) PER2 protein is deleted at residues 512–794 [39], the deletion presumably would not significantly affect the PAS domains, leaving interactions with other PAS proteins including BMAL1 intact. To determine whether the NLD is required for PER2 to bind with BMAL1, we performed coimmunoprecipitation assays in vitro. We found that NLD(−) PER2 can interact as well as intact PER2 with BMAL1. The PAS-A domain of BMAL1 has recently been identified as a critical site for PER2 binding [49]. Therefore, PAS-PAS interaction between PER2 and BMAL1 that presumably occurs in the cytosol might be important for the suppressive effect of PER2 on PAI-1 gene expression. However, we could not determine here whether the interaction between PER2 and BMAL1 is the only mechanism that suppresses CLOCK:BMAL1dependent PAI-1 gene transcription. Biochemical and biophysical differences between BMAL1 and BMAL2 remain unknown. Further studies are needed to understand the molecular mechanism of how the circadian clock component regulates PAI-1 gene expression. We discovered that the diurnal rhythm of exogenous rPer2 gene expression was robust in phase with endogenous mPer2 expression in the hearts of Tg mice, even though the transgene did not include the noncanonical E-box enhancer (CACGTT) that is responsible for circadian transactivation of the Per2 gene in mammals [44]. The mRNA levels of the exogenous Per2 gene fluctuate in a circadian

manner in stable cell lines overexpressing the Per2 gene [51]. Therefore, post-transcriptional regulation such as mRNA stability might be sufficient to drive circadian mRNA expression of the Per2 gene. Circadian expression of clock and clock-controlled genes such as Per1, Per2, Rev-erbα, Cry1, and DBP is positively regulated by CLOCK/ BMAL1 via E-box element(s) in mammals [16,17], whereas the circadian expression of these genes is abolished in Clock mutant mice. A negative feedback loop model in which PER and CRY proteins suppress their own expression by inhibiting CLOCK/BMAL1-dependent transactivation at least in vitro has been postulated [19,25,26]. We found here that the diurnal expression of Per1, Rev-erbα, and DBP was damped in PER2 Tg mice. However, rhythmic expression of endogenous mPer2 was not affected in mice overexpressing PER2. Furthermore, Cry1 expression levels were continuously elevated in PER2 Tg mice, which might have resulted from diminished REV-ERBα expression, because Cry1 expression is regulated not only by CLOCK/ BMAL1 but also directly by the transcriptional repressor, REV-ERBα [52]. Therefore, PER2 seems to regulate Cry1 expression in a bidirectional manner; that is, downregulation is brought about by suppressing CLOCK/BMAL1-dependent transactivation via the E-box element and upregulation results from reducing REV-ERBα levels. The circadian mRNA expression of Per1, Per2, and Cry1 is damped in Per2 mutant mice [28], although PER2 and other PER and CRY proteins suppress the transcription of these genes in vitro [19,25,26]. The present study found that the diurnal expression of endogenous Per2 mRNA was not affected in PER2 Tg mice, suggesting that the circadian clock components are dispensable for rhythmic Per2 expression. Rhythmic Per2 expression in peripheral organs can be driven by both local oscillators and by systemic signals such as body temperature [53] and a robust diurnal rhythm in core body temperature is in fact maintained in PER2 Tg mice [40]. We showed that PER2 suppresses the CLOCK:BMAL1/2-dependent transactivation of mouse PAI-1 gene in vitro. Furthermore, the diurnal expression of PAI-1 mRNA was damped in the hearts of PER2 Tg mice. The attenuating effect of overexpressed PER2 on PAI-1 mRNA expression seemed to result from the direct suppression of E-boxdependent PAI-1 transcription rather than from indirect suppression by diminishing the positive transcription factors (such as CLOCK, NPAS2, BMAL1, BMAL2, RORα, and RORγ) that are involved in circadian transactivation of the PAI-1 gene [13,14], because the expression levels of these transcription factors are identical or upregulated in PER2 Tg mice (Supplemental Fig. 1). Fig. 3 shows that the day/night amplitude of PAI-1 mRNA levels in the hearts of WT mice is 4-fold, whereas Fig. 6 shows that they are b2fold. Cardiac PAI-1 expression seems to be increased as a function of age in normal laboratory mice at 10 and 20 weeks old [54]. This could account for the differences in amplitude between Fig. 3 (about 4 weeks old) and Fig. 6 (16 weeks old). Rev-erbα is involved in the circadian repression of human PAI-1 gene expression by downregulating transcription via two ROREs in the promoter [14]. However, mRNA expression levels of Rev-erbα were reduced in our PER2 Tg mice, suggesting that Rev-erbα is not involved in the suppressive effect of over-expressed PER2 on PAI-1 expression. Although the PAI-1 and Cry1 genes are transcriptionally regulated by both CLOCK/BMAL1 via E-box element(s) and REV-ERBα via RORE(s) [14,52], the effects of PER2 overexpression were antagonistic between the PAI-1 and Cry1 genes in the present study. Therefore, E-boxdependent transcriptional regulation is apparently dominant for circadian PAI-1 expression, whereas RORE-dependent transcriptional regulation is dominant for circadian Cry1 expression. To determine whether PER2 is involved in obesity-induced hypofibrinolysis, we examined PAI-1 gene expression in PER2 Tg mice with diet-induced obesity. Twelve weeks of HFSD feeding significantly increased the BW as well as the plasma cholesterol and insulin levels in both WT and PER2 Tg mice. The insulin-to-glucose

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

ratio (a measure of insulin resistance [47,48]) was significantly increased in both genotypes, although the glucose levels were not affected by obesity. Feeding with HFSD did not affect plasma TG levels in both genotypes, presumably by limiting de novo lipogenesis as reported [55]. Importantly, HFSD did not induce any significant differences in the BW increase, plasma metabolic parameters and glucose metabolism between WT and PER2 Tg mice. These findings suggest that PER2 is not involved in metabolic regulation under dietinduced obesity. Fig. 6 shows that PAI-1 mRNA levels were obviously suppressed under both HFSD and ND in the hearts of PER2 Tg mice, although the mRNA levels in liver and adipose tissues did not differ between the genotypes. Total Per2 (transgenic rPer2 and endogenous mPer2) mRNA levels were remarkably increased in the heart, but not in the liver and adipose tissues of PER2 Tg mice. Tissue-specific suppression of PAI-1 gene expression might have resulted from the different levels of overexpressed PER2 among these tissues in obese PER2 Tg mice. On the other hand, rhythmic expression of BMAL1 was significantly attenuated in the liver and adipose tissues as well as in the hearts of PER2 Tg mice (Supplemental Fig. 4). Therefore, the heart-specific decrease in PAI-1 expression levels might be due not only to the tissue-specific PER2 overexpression levels. To determine whether secondary pathways are involved in the tissue-specific effects on PAI1 gene expression, we examined the mRNA expression levels of TGFβ1, TNF-α, and their receptors in tissues from WT and PER2 Tg mice (Supplemental Fig. 5). However, we could not identify a tissuespecific effect of PER2 overexpression. Levels of PAI-1 mRNA fluctuate in a circadian manner in various tissues such as the heart, liver, lungs, kidneys and adipose tissues [31,56]. The extent to which rhythmic PAI-1 expression depends on the circadian clock apparently differs among tissues. The present findings suggest that circadian clock molecules such as CLOCK, BMAL1/2 and PER2 are more intimately involved in the regulation of cardiac PAI-1 expression than in that of liver and adipose tissues. To determine whether overexpressed PER2 in Tg mice is involved in endogenous PAI-1 expression at the cellular level, we examined PAI-1 mRNA expression in MEF derived from PER2 Tg mice. The role of TGF- β1 in the regulation of PAI-1 expression has been extensively investigated. In turn, PAI-1 is a critical mediator of TGF-β1-induced atherosclerosis [4]. We found here that PAI-1 mRNA expression was significantly suppressed in MEF overexpressing PER2 before and after TGF- β1 stimulation, suggesting that PER2 downregulates PAI-1 gene expression in a cell-autonomous manner. We also found that the TGFβ1-induced mRNA expression of MMP-9 was significantly augmented in MEF from PER2 Tg mice, suggesting that the attenuated PAI-1 induction was not due to impaired TGF-β1 signaling in MEF overexpressing PER2. These observations also suggest that PER2 plays an important role in the regulation of tissue fibrosis and of blood fibrinolysis by regulating PAI-1 and MMP-9 expression. Recent studies have revealed that PER2 has physiological activities that are independent of the circadian clock, such as tumor suppression [20,21], angiogenesis [22], immune function [23], and aortic vascular endothelial function [24]. Furthermore, a dominant familial advanced sleep phase syndrome (FASPS) is linked to post-translational modification resulting from a missense mutation of PER2 [57]. However, whether the association between changes in PER2 function and these observations is mediated by disrupted circadian rhythmicity remains to be confirmed. Further studies will address the physiological association and regulation of PAI-1 expression by PER2. We previously showed that CLOCK is involved in the increased PAI1 production associated with metabolic disorders such as diabetes and obesity [30,31]. However, we could not define whether the Clock mutation directly or indirectly reduces PAI-1 levels by altering metabolic conditions such as lipogenesis and insulin sensitivity. The present findings suggest that PER2, a core component of the circadian clock, suppresses the obesity-induced PAI-1 increase by directly

551

inhibiting PAI-1 gene transcription. We also found that PER2 represses PAI-1 gene transcription in a BMAL1/2-dependent manner. Understanding the molecular mechanism of the obesity-induced plasma PAI-1 increase in relation to the circadian clock should lead to the identification of novel pharmaceutical targets with which to treat obesity or diabetes-induced hemostatic disorders. Acknowledgments We thank Izumi Shibasaki and Hideaki Tanaka (AIST) for animal maintenance and Yumi Ogata (Hokkaido University) for technical assistance. We are grateful to the Kazusa DNA Research Institute for the mPER2 antigen and anti-rPER2 antiserum, and to Dr. Takashi Todo (Kyoto University, Japan) and Dr. Masaaki Ikeda (Saitama Medical University, Japan) for providing the expression plasmids. This project was supported by a Grant-in-Aid for Young Scientists (B) (20770060) to K. Oishi from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan and an operational subsidy from AIST (METI). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.yjmcc.2009.01.001. References [1] De Taeye B, Smith LH, Vaughan DE. Plasminogen activator inhibitor-1: a common denominator in obesity, diabetes and cardiovascular disease. Curr Opin Pharmacol 2005;5:149–54. [2] Alessi MC, Juhan-Vague I. Contribution of PAI-1 in cardiovascular pathology. Arch Mal Coeur Vaiss 2004;97:673–8. [3] Juhan-Vague I, Alessi MC. PAI-1, obesity, insulin resistance and risk of cardiovascular events. Thromb Haemost 1997;78:656–60. [4] Vaughan DE. PAI-1 and atherothrombosis. J Thromb Haemost 2005;3:1879–83. [5] Mutch NJ, Wilson HM, Booth NA. Plasminogen activator inhibitor-1 and haemostasis in obesity. Proc Nutr Soc 2001;60:341–7. [6] Skurk T, Hauner H. Obesity and impaired fibrinolysis: role of adipose production of plasminogen activator inhibitor-1. Int J Obes Relat Metab Disord 2004;28: 1357–64. [7] Manfredini R, Boari B, Smolensky MH, Salmi R, la Cecilia O, Maria Malagoni A, et al. Circadian variation in stroke onset: identical temporal pattern in ischemic and hemorrhagic events. Chronobiol Int 2005;22:417–53. [8] Rana JS, Mukamal KJ, Morgan JP, Muller JE, Mittleman MA. Circadian variation in the onset of myocardial infarction: effect of duration of diabetes. Diabetes 2003; 52:1464–8. [9] Guo YF, Stein PK. Circadian rhythm in the cardiovascular system: chronocardiology. Am Heart J 2003;145:779–86. [10] Yamamoto K, Saito H. A pathological role of increased expression of plasminogen activator inhibitor-1 in human or animal disorders. Int J Hematol 1998;68: 371–85. [11] Andreotti F, Kluft C. Circadian variation of fibrinolytic activity in blood. Chronobiol Int 1991;8:336–51. [12] Maemura K, de la Monte SM, Chin MT, Layne MD, Hsieh CM, Yet SF, et al. CLIF, a novel cycle-like factor, regulates the circadian oscillation of plasminogen activator inhibitor-1 gene expression. J Biol Chem 2000;275:36847–51. [13] Schoenhard JA, Smith LH, Painter CA, Eren M, Johnson CH, Vaughan DE. Regulation of the PAI-1 promoter by circadian clock components: differential activation by BMAL1 and BMAL2. J Mol Cell Cardiol 2003;35:473–81. [14] Wang J, Yin L, Lazar MA. The orphan nuclear receptor Rev-erb alpha regulates circadian expression of plasminogen activator inhibitor type 1. J Biol Chem 2006; 281:33842–8. [15] Oishi K, Shirai H, Ishida N. Identification of the circadian clock-regulated E-box element in the mouse plasminogen activator inhibitor-1 gene. J Thromb Haemost 2007;5:428–31. [16] Reppert SM, Weaver DR. Molecular analysis of mammalian circadian rhythms. Annu Rev Physiol 2001;63:647–76. [17] Reppert SM, Weaver DR. Coordination of circadian timing in mammals. Nature 2002;418:935–41. [18] Vitaterna MH, Selby CP, Todo T, Niwa H, Thompson C, Fruechte EM, et al. Differential regulation of mammalian period genes and circadian rhythmicity by cryptochromes 1 and 2. Proc Natl Acad Sci U S A 1999;96:12114–9. [19] Kume K, Zylka MJ, Sriram S, Shearman LP, Weaver DR, Jin X, et al. mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop. Cell 1999;98:193–205. [20] Fu L, Pelicano H, Liu J, Huang P, Lee C. The circadian gene Period2 plays an important role in tumor suppression and DNA damage response in vivo. Cell 2002; 111:41–50.

552

K. Oishi et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 545–552

[21] Chen-Goodspeed M, Lee CC. Tumor suppression and circadian function. J Biol Rhythms 2007;22:291–8. [22] Koyanagi S, Kuramoto Y, Nakagawa H, Aramaki H, Ohdo S, Soeda S, et al. A molecular mechanism regulating circadian expression of vascular endothelial growth factor in tumor cells. Cancer Res 2003;63:7277–83. [23] Liu J, Malkani G, Shi X, Meyer M, Cunningham-Runddles S, Ma X, et al. The circadian clock Period 2 gene regulates gamma interferon production of NK cells in host response to lipopolysaccharide-induced endotoxic shock. Infect Immun 2006;74:4750–6. [24] Viswambharan H, Carvas JM, Antic V, Marecic A, Jud C, Zaugg CE, et al. Mutation of the circadian clock gene Per2 alters vascular endothelial function. Circulation 2007;115:2188–95. [25] Sangoram AM, Saez L, Antoch MP, Gekakis N, Staknis D, Whiteley A, et al. Mammalian circadian autoregulatory loop: a timeless ortholog and mPer1 interact and negatively regulate CLOCK-BMAL1-induced transcription. Neuron 1998;21: 1101–13. [26] Jin X, Shearman LP, Weaver DR, Zylka MJ, de Vries GJ, Reppert SM. A molecular mechanism regulating rhythmic output from the suprachiasmatic circadian clock. Cell 1999;96:57–68. [27] Zheng B, Larkin DW, Albrecht U, Sun ZS, Sage M, Eichele G, et al. The mPer2 gene encodes a functional component of the mammalian circadian clock. Nature 1999; 400:169–73. [28] Bae K, Jin X, Maywood ES, Hastings MH, Reppert SM, Weaver DR. Differential functions of mPer1, mPer2, and mPer3 in the SCN circadian clock. Neuron 2001; 30:525–36. [29] Ohkura N, Oishi K, Fukushima N, Kasamatsu M, Atsumi GI, Ishida N, et al. Circadian clock molecules CLOCK and CRYs modulate fibrinolytic activity by regulating the PAI-1 gene expression. J Thromb Haemost 2006;4:2478–85. [30] Oishi K, Ohkura N, Amagai N, Ishida N. Involvement of circadian clock gene Clock in diabetes-induced circadian augmentation of plasminogen activator inhibitor-1 (PAI-1) expression in the mouse heart. FEBS Lett 2005;579:3555–9. [31] Oishi K, Ohkura N, Wakabayashi M, Shirai H, Sato K, Matsuda J, et al. CLOCK is involved in obesity-induced disordered fibrinolysis in ob/ob mice by regulating PAI-1 gene expression. J Thromb Haemost 2006;4:1774–80. [32] Oishi K, Miyazaki K, Kadota K, Kikuno R, Nagase T, Atsumi G, et al. Genome-wide expression analysis of mouse liver reveals CLOCK-regulated circadian output genes. J Biol Chem 2003;278:41519–27. [33] Rudic RD, McNamara P, Curtis AM, Boston RC, Panda S, Hogenesch JB, et al. BMAL1 and CLOCK, two essential components of the circadian clock, are involved in glucose homeostasis. PLoS Biol 2004;2:e377. [34] Turek FW, Joshu C, Kohsaka A, Lin E, Ivanova G, McDearmon E, et al. Obesity and metabolic syndrome in circadian Clock mutant mice. Science 2005;308:1043–5. [35] Oishi K, Shirai H, Ishida N. CLOCK is involved in the circadian transactivation of peroxisome-proliferator-activated receptor alpha (PPARalpha) in mice. Biochem J 2005;386:575–81. [36] Oishi K, Atsumi G, Sugiyama S, Kodomari I, Kasamatsu M, Machida K, et al. Disrupted fat absorption attenuates obesity induced by a high-fat diet in Clock mutant mice. FEBS Lett 2006;580:127–30. [37] Kudo T, Tamagawa T, Kawashima M, Mito N, Shibata S. Attenuating effect of clock mutation on triglyceride contents in the ICR mouse liver under a high-fat diet. J Biol Rhythms 2007;22:312–23. [38] Kobayashi Y, Ishikawa T, Hirayama J, Daiyasu H, Kanai S, Toh H, et al. Molecular analysis of zebrafish photolyase/cryptochrome family: two types of cryptochromes present in zebrafish. Genes Cells 2000;5:725–38.

[39] Miyazaki K, Mesaki M, Ishida N. Nuclear entry mechanism of rat PER2 (rPER2): role of rPER2 in nuclear localization of CRY protein. Mol Cell Biol 2001;21:6651–9. [40] Miyazaki K, Wakabayashi M, Chikahisa S, Sei H, Ishida N. PER2 controls circadian periods through nuclear localization in the suprachiasmatic nucleus. Genes Cells 2007;12:1225–34. [41] Hara Y, Shimada K, Kohga H, Ohara O, Koga H. High-throughput production of recombinant antigens for mouse KIAA proteins in Escherichia coli: computational allocation of possible antigenic regions, and construction of expression plasmids of glutathione-S-transferase-fused antigens by an in vitro recombination-assisted method. DNA Res 2003;10:129–36. [42] Todaro GJ, Green H. Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines. J Cell Biol 1963;17:299–313. [43] Lee C, Etchegaray JP, Cagampang FR, Loudon AS, Reppert SM. Posttranslational mechanisms regulate the mammalian circadian clock. Cell 2001;107:855–67. [44] Yoo SH, Ko CH, Lowrey PL, Buhr ED, Song EJ, Chang S, et al. A noncanonical E-box enhancer drives mouse Period2 circadian oscillations in vivo. Proc Natl Acad Sci USA 2005;102:2608–13. [45] Triqueneaux G, Thenot S, Kakizawa T, Antoch MP, Safi R, Takahashi JS, et al. The orphan receptor Rev-erb alpha gene is a target of the circadian clock pacemaker. J Mol Endocrinol 2004;33:585–608. [46] Ripperger JA, Schibler U. Rhythmic CLOCK-BMAL1 binding to multiple E-box motifs drives circadian Dbp transcription and chromatin transitions. Nat Genet 2006;38:369–74. [47] Schreyer SA, Vick C, Lystig TC, Mystkowski P, LeBoeuf RC. LDL receptor but not apolipoprotein E deficiency increases diet-induced obesity and diabetes in mice. Am J Physiol Endocrinol Metab 2002;282:E207–214. [48] Kirk EA, Sagawa ZK, McDonald TO, O'Brien KD, Heinecke JW. Macrophage chemoattractant protein-1 deficiency fails to restrain macrophage infiltration into adipose tissue. Diabetes 2008;57:1254–61. [49] Langmesser S, Tallone T, Bordon A, Rusconi S, Albrecht U. Interaction of circadian clock proteins PER2 and CRY with BMAL1 and CLOCK. BMC Mol Biol 2008;9:41. [50] Lee C, Weaver DR, Reppert SM. Direct association between mouse PERIOD and CKIepsilon is critical for a functioning circadian clock. Mol Cell Biol 2004;24: 584–94. [51] Yamamoto Y, Yagita K, Okamura H. Role of cyclic mPer2 expression in the mammalian cellular clock. Mol Cell Biol 2005;25:1912–21. [52] Liu AC, Tran HG, Zhang EE, Priest AA, Welsh DK, Kay SA. Redundant function of REV-ERB alpha and beta and non-essential role for Bmal1 cycling in transcriptional regulation of intracellular circadian rhythms. PLoS Genet 2008;4:e1000023. [53] Kornmann B, Schaad O, Bujard H, Takahashi JS, Schibler U. System-driven and oscillator-dependent circadian transcription in mice with a conditionally active liver clock. PLoS Biol 2007;5:e34. [54] Sobel BE, Lee YH, Pratley RE, Schneider DJ. Increased plasminogen activator inhibitor type-1 (PAI-1) in the heart as a function of age. Life Sci 2006;79:1600–5. [55] Yazdi M, Ahnmark A, William-Olsson L, Snaith M, Turner N, Osla F, et al. The role of mitochondrial glycerol-3-phosphate acyltransferase-1 in regulating lipid and glucose homeostasis in high-fat diet fed mice. Biochem Biophys Res Commun 2008;369:1065–70. [56] Oishi K, Ohkura N, Kasamatsu M, Fukushima N, Shirai H, Matsuda J, et al. Tissuespecific augmentation of circadian PAI-1 expression in mice with streptozotocininduced diabetes. Thromb Res 2004;114:129–35. [57] Toh KL, Jones CR, He Y, Eide EJ, Hinz WA, Virshup DM, et al. An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome. Science 2001;291:1040–3.